Mechanistic studies of SCS-Pd complexes used in Heck catalysis

Article abstract of DOI:10.1002/adsc.200404270

Air-stable SCS palladacycles that can be used to promote C-C coupling chemistry were studied mechanistically. Using a small library of electronically varied SCS ligands, a collection of palladacycles was synthesized. Kinetic studies showed that these comp

Full text of DOI:10.1002/adsc.200404270

FULL PAPERS
Mechanistic Studies of SCS-Pd Complexes Used in Heck
Catalysis
David E. Bergbreiter,* Philip L. Osburn, Jonathon D. Frels
Department of Chemistry, Texas A&M University, College Station, TX 77842-3012, USA
Fax: (þ1)-979-845-4719, e-mail: Bergbreiter@tamu.edu
Received: August 30, 2004; Accepted: September 27, 2004; Published online: January 11, 2005
Supporting Information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Air-stable SCS palladacycles that can be palladacycle and provided convincing evidence that
ꢀ
used to promote C C coupling chemistry were stud- palladacycles are not the actual catalyst. Poisoning ex-
ied mechanistically. Using a small library of electroni- periments using mercury metal to test for the pres-
cally varied SCS ligands, a collection of palladacycles ence of a Pd colloid were very effective with low mo-
was synthesized. Kinetic studies showed that these lecular weight palladacycles, completely suppressing
complexes all had induction periods, induction peri- Heck chemistry. Similar studies with polymer-bound
ods that were effected by concentration of substrates, palladacycles showed mercury poisoning too. Howev-
products and trace impurities. Hammet correlations er, since so little decomposition of the palladacycle
showed that electronically diverse palladacycles had occurred, the polymer-bound palladacycle could still
identical 1 values, values that suggested that aryl hal- be recycled multiple times. However, mercury pois-
ide electrophilic addition to a Pd species was not the oned subsequent cycles of the experiment too. The
rate-determining step. Phosphine addition experi- conclusion is that SCS palladacycles are actually res-
ments led to increased reactivity of the starting palla- ervoirs of a catalytically active but ill-defined form
dacycles, possibly by trapping an in situ generated of palladium(0).
Pd(0) species. Studies that examined reactivity in bi-
phasic thermomorphic reactions showed residual ac- Keywords: C C coupling, Heck reaction, homogene-
tivity in phases that do not contain polymer-bound ous catalysis, mechanism, metallacycles, palladium
ꢀ
Introduction
cycling catalysts[13,14] and from the initial observations
that these SCS-Pd compounds like their PCP analogues
Pincer complexes of various transition metals have been exhibit improved oxygen and thermal stability com-
of interest in organometallic catalysis and synthesis[1–5] pared to more common phosphine-ligated Pd(0) com-
and in materials chemistry.[6] Palladacycles are the palla- plexes that are used as sources of Pd for Heck cataly-
dium subset of this larger group of metallacycles and sis.[2,6,10,11] We subsequently used these SCS-Pd com-
have been used to promote Pd-catalyzed cross-coupling plexes covalently bound to various soluble polymer sup-
reactions.[7] Palladium pincer compounds are character- ports in Pd-catalyzed cross-coupling reactions.[15–18] In
ized by a metal-carbon bond and two coordinating het- these studies, the SCS-Pd species behave like palladium
eroatoms that chelate the metal in a cyclic fashion catalysts – the covalent polymer-bound palladium spe-
(e.g., 1–3). The most common heteroatoms are P, N, cies are recovered quantitatively and can be reused in
and S. The SCS palladacycles discussed here were first multiple reaction cycles. However, while metal analyses
described by Shaw[8] and they and their N- and P-substi- of product solutions in these systems have generally
tuted analogues have emerged as an interesting, actively failed to detect leaching of Pd, such studies do not distin-
investigated family of organometallic complexes that guish between the SCS-Pd species serving as a molecular
have been used as a palladium source for various catalyt- catalyst or as a precursor of some unknown Pd species
ic reactions.[7,9–12] The mechanistic studies below focus that is the actual active catalyst. Indeed, there is evi-
on a limited subset of these complexes, the so-called dence that suggests that these metallocycles like other
SCS-Pd pincer complexes that have two sulfur atoms bi- and tridentate Pd complexes are catalyst precursors
of a bis(benzyl thioether) chelating the carbon-bound and not catalysts themselves and that Pd(0) species are
palladium. Our interest in these compounds grew out the actual catalysts.[10b,12,19–26] This work includes studies
of our work developing strategies for recovering and re- of SCS palladacycles by our group and by the Jones and
172
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/adsc.200404270
Adv. Synth. Catal. 2005, 347, 172–184

Mechanistic Studies of SCS-Pd Complexes Used in Heck Catalysis
FULL PAPERS
were complete (i.e., if there was no coordination com-
plex 5 present), the S-aryl SCS palladacycles (6, 7, 8, 9,
10, 12, or 13) could be used in Heck chemistry without
formation of any visible palladium black. If the coordi-
nation complex were present as a mixture with the pal-
ladacycle or if the coordination complex itself were
used, palladium black invariably formed. In certain
cases,[10a,22] palladium black also formed from a pallada-
Weck groups[19,20] as well as mechanistic studies of simi- cycle too. While it is difficult to exclude the presence or
lar PCP-Pd complexes.[21] Together these studies pro- involvement of low levels (<10%) of the coordination
vide convincing evidence that these SCS palladacycles complex in any reaction promoted by an SCS-Pd com-
and other palladacycles are neither the catalytically ac- plex, our success in multiple recycling of SCS-Pd com-
tive species nor areversibly formed rest state ofan active plexes that were bound to soluble polymers[15–17] and
catalyst for Heck catalysis but rather are a precursor of our failure to effect similar recycling of S,S-Pd coordina-
another undefined but quite active catalyst.
tion complexes bound to the same polymers suggests
that the catalytic activity seen in any of these reactions
is not due solely to these S,S coordination complexes
or the products of S,S coordination complex decomposi-
tion. Heck chemistry with aryl bromides was confined to
Results and Discussion
We prepared a variety of palladacycles (Figure 1) from a complexes 8, 9, 11, and 13 and was typically seen only
bisbenzyl thioether via an S,S-Pd(II) coordination com- with activated bromoarenes. Table 1 lists the results of
plex (5) [Eq. (1)] using chemistry like that used in earlier these studies. Higher loadings of palladacycles 8, 9 and
work.[8,10]
13 in the presence of 50 mol % of tetrabutylammonium
bromide using K2CO3 as a base were also successful in
complete conversion of bromobenzene to a Heck prod-
uct although the reaction times were long. These results
contrast with the reported activity of PCP-Pd complex-
es[11a] and indicate either that those species generate a
different actual catalyst or that they react in some differ-
ð1Þ ent manner. Reactivity of a palladium species only to-
ward activated aryl halides has been suggested by Du-
The resulting SCS-Pd complexes were shown to pro- Pont to be evidence for Pd(0) colloidal catalysis in reac-
mote Heck reactions of aryl iodides but were less effec- tions promoted by palladium complexes.[26]
tive with aryl bromides. If the palladacycle formation
Figure 1. Electronically diverse SCS palladacycles prepared using metallation of an S,S-Pd(II) coordination complex [Eq. (1)].
Adv. Synth. Catal. 2005, 347, 172–184
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
173

David E. Bergbreiter et al.
FULL PAPERS
Table 1. Heck reactions of aryl bromide with various pallada-
was still speculative prompted us to perform our own in-
vestigations into how SCS palladacycles function in
these catalytic reactions.[19] We hoped to show that the
SCS species are molecular catalysts as it would be
more satisfying to be recycling a molecular catalyst rath-
er than a catalyst precursor. However, we recognized
that while we could recover up to 99.9% of the charged
0.1–1.0 mol % Pd introduced as an SCS-Pd soluble pol-
ymer bound species, very high Pd recovery does not
mean that SCS-Pd compounds themselves are the cata-
lysts. Given the range of electronically varied palladacy-
cles above, we reasoned that these species would have
differing reactivities in the Heck reaction – reactivities
that might allow us to probe the nature of the actual cat-
alyst to determine if these SCS-Pd species are molecular
catalysts, species that reversibly generate a molecular
catalyst, or merely novel repositories of a precursor
form of some unspecified but highly active homeopathic
Pd(0) catalyst.
cycles.[a]
Aryl Bromide
Palladacycle Time [h] Yield [%][b]
4-(NO2)-C6H4-Br
8
9
5.5
4
92
96
10
8
4
4
90
89
4-(CN)-C6H4-Br
9
10
8
9
10
8
5
14
86
81
70
72
4-(CH3CO)-C6H4-Br
26
28
36
51
25
63
C6H5Br
100[c]
100[c]
9
[a]
These reactions were carried out using 0.1 mol % of the
SCS-Pd complex at 1208C in DMF using Et3N as the base.
These yields are isolated yields.
These yields are GC conversions based on aryl bromide
from reactions at 1408C using DMF as the solvent and
K2CO3 as the base in the presence of 50 mol % of Bu4NBr.
[b]
[c]
One approach to study whether the original SCS-Pd
compound was actually a catalyst or merely a catalyst
precursor would be to recover the SCS-Pd compound
and determine if the trifluoroacetate group (or the
chloride group) had exchanged with iodide formed in
Mechanistic Investigations
The development of palladacycles including the SCS these Heck reactions. However, most of our experi-
palladacycles described above and their effective use ments used relatively low amounts (often 0.1 mol %)
ꢀ
in Pd-catalyzed C C bond forming reactions has led to of the SCS palladacycle to promote the reaction, so
a growing interest in the mechanism by which these or- this was not practical. The only case where this issue
ganometallic complexes catalyze coupling reac- was studied was an example where 10 mol % of the
tions.[27–34] This interest was heightened by the sugges- SCS palladacycle 9 was used in a reaction to form methyl
tion that a mechanism involving a Pd(II)-Pd(IV) cycle cinnamate from iodobenzene and methyl acrylate. In
1
could supplement the textbook Heck mechanism in- that instance, examination of the product by H NMR
volving a Pd(0)-Pd(II) cycle. In the most simplistic spectroscopy showed changes (upfield shifts in the aryl
view, substituting the Pd(0) catalyst with a Pd(II) in a methoxy groups and the aryl C-H signals) that suggested
palladacycle like the SCS-Pd species discussed here that the recovered palladacycle was not the starting tri-
would mean that the Pd center is oxidized to Pd(IV) dur- fluoroacetate. However, we did not unambiguously
ing the reaction. Pd(IV) complexes are known and have show that this palladacycle product was in fact the ex-
been extensively studied by Cantyꢁs group,[35] and the pected iodide exchange product.
rather high activity of some members of the palladacycle
To study the effect of electronically different S-aryl
family made the possibility of a different mechanism groups on Heck chemistry, we carried out Heck reac-
seem possible even for the SCS palladacycles described tions under pseudo-first order conditions using 2 mmol
above. However, irreversible or reversible formation of iodobenzene, 20 mmol of methyl acrylate, and
from these palladacycles of a Pd(0) species that is the ac- 20 mmol of Et3N, using 0.001 mmol of an SCS palladacy-
tual active catalyst has never been excluded. Indeed, cle at 1008C. Aliquots were removed at regular intervals
many examples where a Pd(II)-Pd(IV) cycle was sug- (0.5–1 [h]) and analyzed by GC. Reactions were run
gested have been shown to instead involve reduction with palladacycles 6 (S-phenyl), 7 [bis(dimethylamino)],
of the initial Pd(II) complex during the reaction to a 8 [bis(acetamido)], and 12 [bis(benzoic acid)]. Plotting
Pd(0) species. For example, Bohm and Herrmann have the conversion of PhI vs. time showed remarkable simi-
reported a study in which they compared a dimer, phos- larity between the first three complexes (Figure 2). The
phine-based palladacycle with a phosphine-ligated graphs overlap quite well indicating that the electronics
Pd(0) counterpart[33] and concluded that the two cata- of the ligand do not significantly affect the reaction. The
lysts behave similarly enough to preclude aPd(IV) inter- exception to this trend was the reaction using the bis(di-
mediate. The actual catalyst formed from the palladacy- methylamino) palladacycle 7. This reaction converted
cle was not defined.
iodobenzene to product more slowly than the other pal-
The utility of soluble polymer-supported SCS pallada- ladacycles. We speculate that this may be a result of re-
cycles in Heck, Suzuki and Sonogashira reactions and action of the nucleophilic dimethylamino groups with
the fact that mechanistic role of these SCS palladacycles the Heck acceptor which is an alkylating agent, chemis-
174
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 172–184

Mechanistic Studies of SCS-Pd Complexes Used in Heck Catalysis
FULL PAPERS
Figure 2. Conversion vs. time using various SCS palladacy-
Figure 3. Comparison of induction periods using differing
!
!
*
cles. bis(S-phenyl) 6; bis(dimethyl amino) 7; bis(aceta-
*
*
!
amounts of PhI.
1 mmol of PhI;
2 mmol of PhI;
*
mido) 8; bis(benzoic acid) 12.
5 mmol of PhI.
try which could affect the palladacycle or palladacycle riod. A similar experiment studied differing acrylate
decomposition in a complex way. concentrations using three reactions with a constant
The kinetic data shown in Figure 2 also show that 2 mmol of PhI, 20 mmol of Et3N, 0.001 mmol of the pal-
there is an induction period in these reactions. The con- ladacycle 6, and either 2.2 mmol, 5 mmol, or 20 mmol of
version of iodobenzene to product is slow for the first methyl acrylate added. As these reactions were moni-
15–20% of the reaction and then increases. After the tored, the reaction with the smallest excess of methyl
first reaction reached completion, a second portion of acrylate had the shortest induction period while the re-
reagents was added. This second cycle did not exhibit action with the largest excess of methyl acrylate exhibit-
an induction period. The presence of an induction peri- ed the longest induction period (the Figure for this ex-
od when using palladacycles has been reported by oth- periment is provided in the Supporting Information).
ers. Pfaltz and Blackmond investigated a dimeric, nitro- A third set of experiments probed the effect of other
gen-based palladacycle,[33,34] and observed an induction trace species on the reaction rate.
period with this complex and with the dimeric phos-
These experiments were prompted by Pfaltz and
phine palladacycle previously synthesized by Herr- Blackmondꢁs investigations of dimeric, nitrogen-based
mann.[36,37] Similar results were noted by Gladysz in his palladacycles[33,34] where they observed a longer induc-
studies of a recyclable Pd species that was also a Heck tion period in the presence of added water. We per-
catalyst precursor rather than a catalyst.[24] However, formed similar experiments using 2 mmol of PhI,
the presence of an induction period does not provide de- 20 mmol of methyl acrylate, 20 mmol of Et3N and
finitive evidence that the palladacycles are decomposing 0.001 mmol of palladacycle 6 in 10 mL DMF with
as the presence of an induction period could indicate 0.5 mL of water, adventitious water from the DMF sol-
that the complex is altered in some minor fashion to pro- vent, or with predried solvent (drying over molecular
vide the actual catalyst or it could indicate that a species sieves for 24 h). Interestingly, the ꢂwetꢁ reaction had
is generated in the reaction that acts as an accelerant.
the shortest induction period and the ꢂdryꢁ reaction
We performed several experiments to test the effect of had the longest induction period, indicating that the cat-
substrates and other additives on reaction rates and in- alyst formation step is sensitive to the amount of water
duction periods. For example, we varied the amount of present (Figure 4). The reasons for the effect of aryl io-
either iodobenzene or acrylate in the reaction. Three re- dide, acrylate or water on the induction period are not
actions were set up for each experiment. In three vials clear to us. One possibility is that the aryl iodide or acryl-
1 mmol, 2 mmol, and 5 mmol of PhI were each com- ate might be coordinating to the SCS-Pd species, pre-
bined with 20 mmol of methyl acrylate and 20 mmol of venting it from forming the actual more active catalyst.
Et3N. The S-phenyl SCS palladacycle 6 (0.001 mmol) Water then would have to be promoting the formation
was added and the reaction mixtures were heated at of the actual active catalyst.
1008C. Aliquots were removed periodically and ana-
The data above could be explained by decomposition
lyzed by GC. The data from these experiments (Fig- of the SCS palladacycle to for the active catalyst. How-
ure 3) showed that the reaction containing 1 mmol of ever, there are alternative explanations that are consis-
PhI had the shortest induction period while the reaction tent with these data that do not require the palladacycle
containing 5 mmol of PhI had the longest induction pe- to decompose to form another unspecified catalyst. For
Adv. Synth. Catal. 2005, 347, 172–184
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
175

David E. Bergbreiter et al.
FULL PAPERS
bond would prevent the subsequent elimination of hal-
ide that is part of Shawꢁs mechanism. They did not detect
fluoride-containing products and they also concluded
that fluoride does not attack the olefin during the reac-
tion.
We also considered the possibility that acid generated
during the reaction might in some way be accelerating
the reactions. In reactions like those described above,
0, 10, or 20 mol % of concentrated HCl was added
(HCl was chosen instead of HI because HI has a tenden-
cy to liberate I2 and it was unclear how the presence of
iodine might affect the reaction). Like the reactions in
which Bu4NI was added, the reactions containing added
HCl showed longer induction periods compared with
the reaction containing no additives. These experiments
Figure 4. Comparison of induction periods with varying with added iodide or protons suggest the species gener-
*
*
!
amounts of water. ꢂwetꢁ; ꢂnormalꢁ; ꢂdryꢁ.
ated during the Heck reaction are probably not assisting
the reaction and that the nucleophilic assistance describ-
ed by Shaw is not part of the catalytic cycle. Why the in-
example, Shaw[31,32] envisioned a possible pathway duction period is affected by the presence or absence of
where the olefin first coordinates to the Pd center and additives is not clear. However, the notion that the SCS-
a nucleophile then attacks the olefin, forming a palladi- Pd species is not the actual catalyst but rather is decom-
um-alkyl complex. Shaw proposed that the increased posing under Heck reaction conditions to form the ac-
electron density of this newly formed complex facilitates tual catalyst is the simplest explanation of these experi-
the oxidative addition of the aryl halide. This mecha- ments.
nism provides possible explanations for both the ob-
served induction period and the accelerating effect of
water. The induction period could be the product of Using Substrates as Mechanistic Probes
the generation of iodide, a potential nucleophile. In
the early stages of the reaction, the concentration of io- We studied the reactivity of different aryl iodides in
dide is low. As the reaction progresses, the concentration Heck reactions promoted by different SCS palladacy-
of iodide increases, facilitating the formation of the pal- cles in the hope that Hammet plots would provide useful
ladium-alkyl complex until a limiting concentration of information. These competition experiments were per-
iodine is reached. Water could also serve as a nucleo- formed in two ways.[39] In the first method, a substituted
phile, thereby explaining the shortened induction period iodoarene like iodoacetophenone and iodobenzene
in the presence of water.
(2 mmol each) were combined in a flask containing
We tested this idea using a set of experiments where 20 mmol of methyl acrylate and 20 mmol of base. The re-
tetrabutylammonium iodide was added. If the chemistry action was carried out using 0.002 mmol of a palladacy-
proceeded according to Shawꢁs mechanism, the added cle at 1008C. Periodically, aliquots were removed from
iodide should eliminate or greatly reduce the induction the reaction and analyzed by GC. The fraction remain-
period. Three reactions were set up using 2 mmol of PhI, ing of each iodoarene was calculated at each time inter-
20 mmol of methyl acrylate, and 20 mmol of Et3N in the val. The reaction was run until the one of the iodoarenes
presence of 0.001 mmol of S-phenyl palladacycle 6. The was completely consumed. Then the log of the calculat-
first reaction was performed without additives, the sec- ed fraction of iodoarene at each time interval was plot-
ond with 10 mol % (relative to the aryl iodide) of ted against the log of the fraction of iodobenzene at
Bu4NI, and third with 20 mol % of Bu4NI. The reactions the same time interval, producing a straight line (Fig-
with added Bu4NI exhibited increased induction periods ure 5). The slope of that line is the relative reactivity
(see Supporting Information Figure 2). There was little of the iodoarene vs. iodobenzene. Separate reactions
difference in the induction periods for the two reactions were performed using iodoacetophenone, iodotoluene,
containing differing amounts of Bu4NI. All three reac- and iodoanisole and relative reactivities were deter-
tions eventually went to completion. The extended in- mined in each reaction.
duction period in the presence of added iodide was in-
The second method combined all four iodoarenes in
consistent with Shawꢁs mechanism. Crabtreeꢁs group the same reaction. The amount of each iodoarene was
has reported a similar experiment in which they added reduced to 1 mmol to maintain a total of 4 mmol of io-
NaF to a Heck reaction catalyzed by a bis-carbene, pin- doarene in the reaction. The other parameters were
cer-type palladacycle.[38] They reasoned that if fluoride identical as those described above. In this method, the
ꢀ
nucleophilically attacked the olefin, the strong C F reaction was allowed to continue until the iodobenzene
176
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 172–184

Mechanistic Studies of SCS-Pd Complexes Used in Heck Catalysis
FULL PAPERS
Table 2. Reactivity ratios obtained with various SCS pallada-
cycles.[a]
Aryl iodide
7
6
11
iodoacetophenone
iodobenzene
iodotoluene
11.8
1
0.67
0.56
12.4
1
0.66
0.59
11.5
1
0.69
0.62
iodoanisole
[a]
In these experiments, 4 mmol of an iodoarene mixture was
allowed to react with 20 mmol of methyl acrylate and
20 mmol of triethylamine base in the presence of
0.002 mmol of the indicated SCS palladacycle at 1008C.
Aliquots were periodically removed from the reaction
and analyzed by gas chromatography.
Figure 5. Plot derived from competition experiment between
iodobenzene and iodoacetophenone.
had been consumed. The fractions remaining for each
iodoarene were calculated and the data for iodoarenes
were plotted against that for iodobenzene. This method
provided equivalent results as those obtained by the pre-
vious method but shortened the amount of time necessa-
ry to collect all the data.
The reactivity ratios measured in these experiments
were used to construct Hammet plots using both s and Pd(0) catalysts, suggesting that the palladacycle cata-
s– substituent constants. Good linearity was achieved lysts are not the same. The values obtained here also
when using s values (r2 ¼0.96) but a better fit was ob- do not seem to be high enough to indicate that the oxida-
tained using s– values (r2 ¼0.99). This small difference tive addition of the aryl halide is the rate-determining
may indicate that a direct resonance interaction exists step. Milsteinꢁs group studied the addition of aryl chlor-
between the para substituent and the reactive center. ides to a Pd(0) catalyst ligated with bidentate phos-
The slope of both graphs was positive, indicating that phines.[40] They performed competition experiments
electron-withdrawing groups facilitate the reaction. with substituted aryl chlorides and obtained r values
The value of r for the plot derived using s values was of approximately 5, so showing that the oxidative addi-
1.86 while a value of 1.2 was obtained using s– values. tion of the aryl chloride is rate-determining. The obser-
In a similar way, reactivity ratios were determined using vation that palladacycles do not exhibit a similar r value
bis(dimethylamino) SCS 7 and bis(tert-butyl) SCS 11 casts doubt on the oxidative addition step being rate de-
(Table 2). The reaction constants for Heck reactions us- termining.
ing these Pd species were identical to the reaction con-
stant measured for reactions using 6.
These 1 values are comparable to those obtained by Added Phosphine Experiments
other groups. Milstein[11a] reported a r value of 1.39 vs.
s values for the reaction of aryl iodides catalyzed by tri- To test the possibility that the organometallic com-
dentate PCP palladacycle 14. Herrmann[27] reported a r pounds simply act as ꢂreservoirsꢁ of Pd(0) species like
value of 1.58 vs. s– values with aryl bromides using bi- those present in phosphine ligated Pd(0) catalysts, we
dentate phosphapalladacycle 15. In the same article, carried out a Heck reaction using iodobenzene and
he reported that a tri-o-tolylphosphine ligated Pd(0) cat- 3,5-dimethylbromobenzene in the presence of methyl
alyst 16 produced a r value of 1.01 (vs. s– values) for the acrylate and triethylamine.[41,42] S-Phenyl palladacycle
same substrates. Using bidentate sulfur-based pallada- 6 was used as the Pd source and 2 equivs. (relative to
cycle 17, Dupont[10e] reported that in a competition ex- the palladacycle) of triphenylphosphine were also add-
periment iodoacetophenone, iodobenzene, and iodoa- ed. If a Pd(0) species were formed, we reasoned that
nisole are converted to their corresponding cinnamate the phosphine could coordinate to the Pd(0), forming
products in a ratio of 7:2:1, respectively. This result a catalyst that is active for the coupling of aryl bromides
translates to a Hammet plot with a r of 1.13 vs. s values. and acrylates. Since the SCS complex used does not con-
Our results are thus consistent with similar catalysts. The vert 3,5-dimethylbromobenzene to a Heck product, any
r values obtained are different from phosphine-ligated observed conversion of the aryl bromide would be evi-
Adv. Synth. Catal. 2005, 347, 172–184
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
177

David E. Bergbreiter et al.
FULL PAPERS
dence for phosphine-ligated Pd(0) formation due to de-
composition of the starting SCS palladacycle.
The reaction was performed under nitrogen using
0.5 mmol of each aryl halide, 1.1 mmol of methyl acryl-
ate in DMFat 1108C in the presence of SCS palladacycle
6 (1 mol % relative to total aryl halide), PPh3 (2 mol %)
and Et3N (1.1 mmol) [Eq. (2)]. The reaction was moni-
tored by GC. As the reaction was followed, the iodoben- and ultimately the metal center is transferred from the
zene was consumed but the 3,5-dimethylbromobenzene NCN moiety to the PCP ligand.
remained untouched. The bromoarene was not convert-
Since the reaction of an added phosphine with the SCS
ed to the cinnamate product over the course of 94 h. palladacycles would confuse the simple suggestion that
However, using the S-t-Bu SCS palladacycle 11, we did phosphines like Ph3P or (t-Bu)3P trap an SCS palladacy-
observe some conversion of the aryl bromide. Approxi- cle Pd(0) decomposition product, we dissolved the S-
mately 15% of the 3,5-dimethylbromobenzene was con- phenyl SCS palladacycle 6 in DMSO-d6 along with 2
sumed in 16 h when 11 was used as the Pd source. The equivs. of PPh3 and studied the resulting mixture by
formation of some 3,5-dimethylbromobenzene-derived NMR spectroscopy. A 1H NMR spectrum of this sample
product using a phosphine ligand with SCS palladacycle showed that the signal corresponding to the benzyl pro-
11 is consistent with earlier observations that this palla- tons had moved from d¼4.8 to d¼4.1 ppm. The signal
dacycle is not stable under the conditions of Heck cata- assigned to the amide N-H at d¼9.85 was shifted upfield
lysis.[10a,25]
to d¼9.4, and the acetamido CH3 signal moved from
d¼2.0 to d¼1.95 ppm. Two peaks were seen in the 31
P
NMR spectrum – one at d¼26.6 and one at d¼23.1.
One peak is likely triphenylphosphine oxide. The sec-
ond peakꢁs chemical shift is similar to the shifts observed
by Stang109 for the phosphine aryl-palladium complexes
18 and 19. Given that PPh3 evidently reacts with the pal-
ladacycle 6, we cannot unambiguously conclude that the
phosphine addition experiments really do show that
Pd(0) formation is occurring. It is instead equally possi-
ð2Þ
Although conversion of 3,5-dimethylbromobenzene ble that added phosphine is displacing the sulfur atoms
was not observed in the experiment using S-phenyl pal- and coordinating to the Pd center [Eq. (3)] to form a
ladacycle 6, it is possible that the amount of Pd(0) pseudo-PCP catalyst that reductively eliminates to unin-
formed (and consequently the concentration of a PPh3 tentionally form a phosphine-ligated Pd(0) species that
ligated Pd species) was too low to affect detectable con- in turn converts 3,5-dimethylbromobenzene to Heck
version of 3,5-dimethylbromobenzene in a reasonable products.
amount of time. We therefore repeated this experiment
using the bulky ligand, tri(tert-butyl)phosphine. Palladi-
um catalysts using this phosphine are active enough to
ð3Þ
convert aryl chlorides even at room temperature,[43] so
we reasoned that this phosphine would be more sensi-
tive test of Pd(0) formation. Using the previously de-
fined conditions and S-phenyl palladacycle 6, P(t-Bu)3 Although the above experiments all suggest that SCS
was substituted for PPh3. In contrast to the reaction palladacycles decompose to form some other Pd species
above, this experiment did show a 10% conversion of that is the actual catalyst in cross-coupling reactions pro-
3,5-dimethylbromobenzene to a Heck product in 14 h.
moted by SCS-Pd palladacycles, each experiment has an
While these experiments with added phosphine li- alternative explanation. Since we have not been able to
gands provide additional support to the notion that unambiguously detect Pd colloids or identify some other
SCS-Pd complexes are catalyst precursors, the observa- species as the actual catalyst,[25] we sought to develop
tions above are complicated by the fact that added phos- other experiments that might more definitively show
phine is not an innocent spectator in the presence of a that SCS palladacycles on polymer supports are not
palladacycle. Shawꢁs group had earlier noted that PPh3 the actual catalysts. These experiments involved using
coordinates to the palladium center of an SCS com- biphasic studies of polymer-supported SCS palladacy-
plex.[8] van Koten has also shown that unmetallated cles where the actual catalyst has a lower phase selective
PCP ligands can react with metallated NCN complexes solubility than the polymer-bound SCS palladacycle cat-
in a trans-cyclometallation reaction.[8,44] During this re- alyst precursor or where the phase separability of the
action, the nitrogen ligands of the NCN complexes are polymer-bound SCS palladacycle from heterogeneous
displaced by the phosphorus atoms of the PCP ligand, Pd(0) colloid traps [Hg(0)] allowed us to show that recy-
178
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 172–184

Mechanistic Studies of SCS-Pd Complexes Used in Heck Catalysis
FULL PAPERS
cled palladacycles generate species whose activity is ble in the heptane-rich phase at 258C. In contrast, PNI-
poisoned by elemental mercury. PAM polymers are exclusively (>10,000:1) soluble in
These experiments used the polymer-bound ligands the DMA-rich phase at 258C.[47]
20 and 22 and the polymer-bound SCS palladacycles
In the first experiment, the PNIPAM-SCS-Pd species
21 and 23. Such species have been used by us previously 21 (0.2 mol %) was used in a Heck reaction of iodoben-
in Heck catalysis[15–18] and are accessible by chemistry zene and methyl acrylate (1008C, 6 h, Et3N as the base).
like that shown in Eq. (4). The studies below were car- After the iodoarene was consumed, the hot single phase
ried out in a thermomorphic system of heptane and reaction mixture was cooled to form a biphasic mixture.
90% DMA.[45,46]
The top heptane-rich phase was shown to have
<0.1 ppm Pd (ICP analysis). Furthermore, addition of
fresh substrates to this phase and heating led to no
Heck chemistry. A second experiment was then carried
out with the same substrates using PNODAM-SCS-Pd
23 (0.2 mol %) as the Pd source. Again, Heck chemistry
occurs in the thermomorphic mixture (6 h for complete
conversion of iodobenzene). On cooling, the polar
DMA-rich phase was isolated and analyzed for Pd.
Again, no Pd (<0.1 ppm) was detected by ICP analysis.
However, addition of fresh substrates to this phase and
heating did lead to Heck catalysis. Such activity was
also observed for the polar phase of a second cycle
when the heptane soluble 23 was reused in a thermomor-
phic system. While the Heck reaction seen with these
polar phases was slower (14 h for complete conversion
of iodobenzene), the observation of any activity in a
ð4Þ
Aqueous DMA was used to avoid solubility problems phase that contains no polymer and hence no palladacy-
observed with the PNIPAM-SCS system. In this system, cle is unambiguous evidence that Pd dissociates from the
PNODAM polymers are exclusively (>10,000:1) solu- polymer and that the palladacycle is not required for cat-
alysis.
We also carried out one set of experiments using the
polymer-bound palladacycle 23 and the polar phase
soluble SCS ligand 20 to qualitatively examine the
hypothesis that these palladacycles reversibly form
some coordination complex during the reaction. We rea-
ꢀ
soned that if the Pd C bond is reversibly cleaved during
the reaction, reinsertion of the Pd to reform a palladacy-
cle could distribute the palladium between the two pol-
ymer bound ligands (Figure 6). If this exchange oc-
curred, we would then expect to observe higher catalytic
activity in the aqueous DMA layer containing PNI-
PAM-bound SCS ligand than in a DMA layer that con-
tained no polymer. The result of this experiment was
that the polar phase had activity but that the activity
Figure 6. Potential scrambling pathway between PNODAM and PNIPAM-bound SCS ligands.
Adv. Synth. Catal. 2005, 347, 172–184
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
179

David E. Bergbreiter et al.
FULL PAPERS
was the same as that of a polar phase where no polymer mediate. A definitive experiment for decomposition of
was present. the palladacycle to some other active catalyst was the re-
The species that is generated in these reactions could sidual catalytic activity in the separated phase of a ther-
be a palladium colloid. To examine this issue, we fol- momorphic system. The observed catalytic activity in a
lowed the lead of others who have used mercury metal solution that was separated from the polymer-bound
as a trap for a colloidal metal species.[20,21,48–50] Such ex- SCS palladacycle even with recycled polymer-bound
periments have successfully implicated Pd colloids in palladacycles is a strong indication that some other
Heck couplings using PCP-Pd complexes and more re- non-supported palladium catalyst is formed in situ.
cently in SCS-Pd complexes. Our results agreed with The work by other groups on other palladacycles where
these other findings.
Hg poisoning is seen and in our work where Hg retarda-
Several experiments were performed. One reaction tion of a polymer-bound palladacycleꢁs reactivity impli-
used iodobenzene, methyl acrylate, 0.1 mol % pallada- cates the presence of Pd(0) colloids or nanoparticles as
cycle 6, and an excess (with respect to the palladacycle) the actual catalytic species.
of Hg. The reaction mixture was heated at 1008C and
The conclusion that some palladium species other
monitored but no conversion of the starting materials than the palladacycle (e.g., a palladium nanoparticle or
to product was observed. A similar reaction without colloid) is the actual catalytic species is something of a
Hg did form products. This second Hg-free reaction disappointment as it indicates that our earlier work recy-
however, stopped if Hg were added. More complex re- cled a catalyst precursor and not the catalyst itself. That
sults were seen in Heck chemistry with the polymer- having been said, it must be remembered that these pal-
bound palladacycle 23. In a Heck reaction of methyl ladacycles are active for the Heck reaction at fairly low
acrylate and iodobenzene with 0.1 mol % of 23 without loadings. Moreover, these SCS palladacycles have been
Hg, the iodobenzene and methyl acrylate were smoothly successfully used in recycling strategies on different pol-
and completely converted to methyl cinnamate. Howev- ymer supports by a variety of groups and Pd loss in these
er, in the reaction containing Hg, the reaction reached experiments is inconsequential. Based on this recycling
20% conversion in 18 h but did not continue past this work and palladium analysis, the amount decomposition
point. The polymer might be shielding the colloidal pal- of the palladacycle is low. Less than 1% of the original
ladium for a short time, allowing some conversion of the palladacycle is transformed during the reaction. For a
starting materials. In this case, the polymeric species palladacycle loading of 0.2 mol % this translates to a
could still be recovered by heptane extraction and the turnover number of 50,000 or better for the actual cata-
recovered 23 could be used in a second cycle, promoting lyst.
Heck chemistry of these same substrates. The reactivity
of 23 in this second cycle was qualitatively unchanged
from its reactivity in an Hg-free first cycle. Reactivity
of the recycled 23 was, however, again poisoned on the
addition of Hg. These results, like the results throughout
these studies are again consistent with these palladacy-
Experimental Section
cles being relatively stable species that under Heck reac-
tion conditions generate an undefined, possible colloi-
dal Pd, catalyst.
General
Reagents and solvents were obtained from commercial sources
and were generally used without further purification. Gas
chromatographic analyses were performed on a Shimadzu in-
strument equipped with a 15-m SPB-5 [poly(5%-diphenyl-
95%-dimethylsiloxane)] normal phase fused silica capillary
Conclusions
1
Electronically varied tridentate SCS ligands and palla-
dacycles were synthesized and were active in promoting
the Heck reaction of aryl iodides but were less active in
Heck chemistry of aryl bromides. Mechanistic investiga-
tions using these palladacycles showed that electronical-
ly different palladacycles had the same reactivity. Reac-
tions with these palladacycles exhibited inductions peri-
ods that were dependent in a complex way on the start-
ing material concentration, adventitious water and oth-
er additives. Competition experiments between various
aryl iodides yielded Hammet plots that remained consis-
tent from palladacycle to palladacycle. Phosphine addi-
tion that led to reactivity uncharacteristic of the pallada-
column (0.53 ID). H NMR spectra were recorded on Varian
spectrometers at 300 or 500 MHz. Chemical shifts are reported
in ppm using hexamethyldisiloxane (HMDS, 0.055 ppm) as the
internal standard. 13C NMR spectra were recorded at 75 MHz
or 125 MHz with CDCl3 (77.0 ppm) or DMSO-d6 (39.5 ppm) as
the internal reference. Solid state NMR spectra of 23 were ob-
tained using a Bruker MSL-300 MHz NMR spectrometer. In-
frared spectra were recorded as thin films between NaCl plates
or as pressed KBr pellets using a Mattson Galaxy 4021 FT-IR
spectrometer. Metal analyses were performed using a Beck-
man Spectra Span VI operating with the AdaM VI software
package. Melting points were determined with a Thomas-Hoo-
ver Unimelt capillary melting point apparatus and were uncor-
rected. S,S ligands and SCS-Pd complexes not explicitly descri-
cycle was consistent with the trapping of a Pd(0) inter- bed below were prepared using literature procedures.[10a,15,16]
180
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 172–184

Mechanistic Studies of SCS-Pd Complexes Used in Heck Catalysis
FULL PAPERS
PNODAM-NASI
N-Acetyl-3,5-bis((p-
dimethylaminophenyl)thiomethyl)aniline
Benzene (50 mL) was bubbled through with nitrogen for
30 min. Octadecylacrylamide (4.5 g, 13.9 mmol) and N-acryl-
oxysuccinimide (0.24 g, 1.39 mmol) were added. The reaction
mixture was warmed to 808C and a solution of AIBN
(12.6 mg, 0.077 mmol) in 2 mL of benzene was added. The mix-
ture was stirred at 808C under nitrogen for 36 h. The solvent
was removed and the residue taken up in 30 mL of chloroform
and precipitated into 150 mL of cold methanol to give a white
powder; yield: 4.1 g (87%). IR (KBr): n¼1812, 1783, 1737,
1654, 1542 cmꢀ1; 1H NMR (CDCl3): d¼0.8 (t, 30H), 1.2 (br s,
395H), 1.5 (br s, 27H), 2.8 (br s, 4H), 3.1 (br s, 20H).
A 200-mL flask with attached condenser was purged with ni-
trogen. Acetone (50 mL) was bubbled through with nitrogen
for 15 min. Dimethylaminothiophenol (590 mg, 3.85 mmol),
K2CO3 (604 mg, 4.38 mmol), and N-acetyl-3,5-bis(benzyl
chloride)[10a] (407 mg, 1.75 mmol) were combined in the 200-
mL flask and dissolved in 100 mL of acetone. The reaction mix-
ture was protected from light and refluxed for 48 h. The reac-
tion mixture was filtered and the solvent removed. The residue
was dissolved in chloroform and washed with water and brine.
The solvent was evaporated and the residue purified by chro-
matography on silica gel (ethyl acetate:CH2Cl2, 1:9) to afford
PNIPAM-SCS (20) and PNIPAM-SCS-Pd (21) were pre-
pared following a literature procedure.[15]
a
white solid; yield: 574 mg (70%); mp 115–116.58C.
1H NMR (DMSO-d6): d¼1.99 (s, 3H), 2.83 (s, 12H), 3.90 (s,
4H), 6.60 (dd, 4H), 6.78 (s, 1H), 7.17 (dd, 4H), 7.39 (s, 2H),
9.84 (s, 1H); 13C NMR (DMSO-d6): d¼24.0, 39.9, 40.4, 112.7,
117.8, 119.8, 124.1, 133.2, 138.6, 139.2, 149.7, 168.2; IR (KBr):
PNODAM-SCS (22)
n¼3292, 1664, 1594 cmꢀ1
;
HRMS: m/z calcd. for
In 30 mL of tert-butanol, PNODAM-NASI 21 (300 mg,
0.089 mmol) was dissolved. Separately, amine-terminated
SCS ligand 51[64] (50 mg, 0.089 mmol) was dissolved in 10 mL
of tert-butanol with 0.5 mL of Et3N. This solution was transfer-
red to the polymer solution and the reaction mixture was stir-
red at 558C for 4 h. The solvent was removed under reduced
pressure. The residue was taken up in 6 mL of chloroform
and precipitated into 40 mL of methanol to afford the polymer
as a white powder; 310 mg (93%). 1H NMR (CDCl3): d¼0.9 (t,
30H), 1.2–1.5 (br m), 3.1 (br s, 20H) 4.0 (br s, 3H), 6.9 (br s,
1.5H), 7.1–7.3 (br m, 8H), 7.5 (br s, 1.5H); 13C NMR (CDCl3)
d¼175.4, 139.0, 138.3, 136.3, 129.6, 128.7, 126.1, 124.5, 119.0,
42.4, 39.7, 38.8, 31.8, 29.7, 29.4, 29.3, 28.0, 27.2, 22.6, 13.9.[16]
C24H21O5NS2 [MþHþ]: 466.1987; found: 466.1984.
N-Acetyl-3,5-bis((p-
acetamidophenyl)thiomethyl)aniline
N-Acetyl-3,5-bis(benzyl chloride) (0.5 g, 2.15 mmol), p-acet-
amidothiophenol (880 mg, 5.25 mmol), and K2CO3 (725 mg,
5.25 mmol) were combined in a 100-mL flask. Acetone
(50 mL) was added and a reflux condenser attached. The flask
was covered infoil, flushed with argon and the reaction mixture
brought to reflux. After 15 h, the mixture was filtered and the
solvent removed under reduced pressure. The resulting solid
was broken up, placed on a filter, washed with portions of wa-
ter, and dried on the aspirator to afford a white powder; yield:
936 mg (88%); mp 163–1658C. 1H NMR (DMSO-d6): d¼
2.0(overlapping singlets, 9H), 4.08(s, 4H), 6.91(s, 1H), 7.22
(dd, 4H), 7.48 (s, 2H), 7.53 (dd, 1H), 9.90 (s, 1H), 9.95 (s, 2H);
13C NMR (DMSO-d6): d¼23.8, 117.9, 119.4, 123.9, 129.0,
130.0, 137.8, 138.1, 139.3, 168.1; IR (KBr): n¼3299, 2923,
1664, 1598 cmꢀ1; HRMS: (m/z) calcd. for C24H21O5NS2 [Mþ
Hþ]: 494.1572; found: 494.1660.
PNODAM-SCS-Pd (23)
PNODAM-SCS 22 (256 mg, 0.068 mmol) was dissolved in
20 mL of THF. To this was added a solution of Pd(NCPh)2Cl2
(26 mg, 0.068 mmol) in 5 mL of THF. The solution was stirred
at room temperature for 1 h, then brought to reflux for 24 h.
The solvent was removed on a rotary evaporator and the resi-
due dissolved in 6 mL of CHCl3 and precipitated into 40 mL of
methanol affording a brownish powder; yield: 246 mg (94%).
1H NMR (CDCl3): d¼0.9 (t, 30H), 1.1–1.5 (br m), 3.1 (br s,
31H), 4.5 (br s, 2.4H), 7.3 (br s, 2.5H), 7.8 (br s, 1H);
13C NMR (solid state): d¼152.2 (ipso carbon).[16]
N-Acetyl-3,5-bis((benzoxazole)thiomethyl)aniline
Thermomorphic Heck Reaction
In an argon purged flask, N-acetyl-3,5-bis(benzyl chloride)
(0.5 g, 2.15 mmol) was dissolved in 40 mL of acetone. To this
was added K2CO3 (743 mg, 5.38 mmol) and a solution of mer-
captobenzoxazole (720 mg, 4.74 mmol) in 10 mL of acetone.
The reaction mixture was protected from light and refluxed
overnight. The mixture was filtered and the solvent removed
under reduced pressure. The residue was taken up in chloro-
form, washed with water, dried and the solvent removed on
the rotary evaporator; yield: 659 mg (66%); mp 164–1658C.
1H NMR (DMSO-d6): d¼10.02 (s, 1H), 7.65 (s, 2H), 7.60 (m,
4H), 7.30 (m, 5H), 4.59 (s, 4H), 2.00 (s, 3H); 13C NMR
(DMSO-d6): d¼168.3, 163.5, 151.2, 141.1, 139.7, 137.2, 124.5,
124.2, 124.0, 118.5, 118.2, 110.0, 35.4, 23.8.
In a screw cap vial, 2 mL of a heptane solution of PNODAM-5
(7.5 mg/mL, 0.004 mmol Pd total) was diluted to 5 mL with
heptane. Iodobenzene (2 mmol), acrylic acid (2.5 mmol), and
Et3N (6 mmol) were dissolved in 10 mL of DMA and added
to the vial. The reaction mixture was heated at ca. 1008C. After
the reaction, the vial was cooled and the layers allowed to sep-
arate. The lower DMA phase was removed, poured into water
and acidified with HCl to precipitate the product. The product
was then filtered and dried. Fresh reactants in DMAwere then
added for the next cycle. The same procedure was used with
PNIPAM-SCS-Pd complex, the difference being that the PNI-
PAM-bound complex was recovered in the polar phase.
Adv. Synth. Catal. 2005, 347, 172–184
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
181

David E. Bergbreiter et al.
FULL PAPERS
N-Acetyl-3,5-bis((p-carboxyphenyl)thiomethyl)aniline Palladation of N-acetyl-3,5-bis((p-
carboxyphenyl)thiomethyl)aniline
A 100-mL flask and condenser were purged with argon. The
flask was charged with N-acetyl-3,5-bis(benzyl chloride)
(0.5 g, 2.15 mmol), 4-mercaptobenzoic acid (730 mg,
4.73 mmol) and K2CO3 (654 mg, 4.73 mmol). Ethanol
(20 mL) and water (10 mL) were added, the flask was covered
with foil and the reaction mixture refluxed for 2 h. The mixture
was concentrated under reduced pressure, acidified with 5 mL
of 10% HCl, and 150 mL of water were added. The product was
The bis(benzoic acid) SCS ligand described above (100 mg,
0.21 mmol) was dissolved in 10 mL of glacial acetic acid.
Pd(OAc)2 was dissolved in 10 mL of glacial acetic acid and add-
ed to the ligand solution dropwise from a pipette. A precipitate
formed within a few minutes. The suspension was stirred over-
night at room temperature and then refluxed for an additional
60 h. The reaction mixture was cooled and filtered. The filtrate
allowed to precipitate as a white solid (2 h) and was then isolat- was evaporated to dryness and the resulting yellow solid sus-
ed by filtration; yield: 800 mg (80%); mp 2128C (dec). 1H NMR pended in 10 mL of acetonitrile and refluxed overnight. The re-
action mixture was cooled then filtered to afford 12 as a yellow
(DMSO-d6): d¼1.99 (s, 3H), 4.24 (s, 4H), 7.15 (s, 1H), 7.38 (dd,
solid; yield: 51 mg (38%); mp>2708C. 1H NMR (DMSO-d6):
4H), 7.57 (s, 2H), 7.81 (dd, 1H), 9.95 (s, 1H), 12.9 (s, 2H);
13C NMR (DMSO-d6): d¼24.0, 35.4, 118.2, 124.2, 126.4,
127.4, 129.7, 137.5, 139.7, 143.0, 166.9, 168.4; IR (KBr): n¼
1687, 1592 cmꢀ1; HRMS: m/z calcd. for C24H21O5NS2 [Mþ
Hþ]: 468.0939; found: 468.0959.
d¼d 2.0 (s, 3H), 4.82 (s, 4H), 7.26 (s, 2H), 7.95–7.8 (overlap-
ping dd, 8H), 9.82 (s, 1H); 13C NMR (DMSO-d6): d¼23.8,
38.1, 49.1, 113.2, 130.0, 136.4, 149.4, 168.0; IR (KBr): n¼
3426, 1700, 1592, 1537 cmꢀ1
; HRMS: m/z calcd. for
C26H23O7NS2Pd [MþHþ þH2O]: 650.0177; found: 649.9864.
General Procedure for Heck Reactions using Aryl
Halides and SCS-Pd Complexes
Palladation of N-Acetyl-3,5-bis((p-
A screwcap vial was charged with aryl bromide (5.0 mmol),
methyl acrylate (7.5 mmol), triethylamine (7.5 mmol) and
DMF (10 mL). tert-Butylbenzene (1.0 mmol) was then added
as an internal standard. The catalyst (0.1 mol %) was added
as a solution in DMF, and the reaction solution was heated to
1208C with an oil bath. The reactions were monitored by
GC. After completion, the reaction solution was poured into
chilled water (150 mL) and the product isolated by filtration.
dimethylaminophenyl)thiomethyl)aniline
In a 50-mL round-bottomed flask, the dimethylamino S,S li-
gand prepared above (100 mg, 0.215 mmol) was dissolved in
10 mL of acetone. Pd(TFA)2 (75 mg, 0.225 mmol) was dis-
solved in 10 mL of acetone and added to the reaction mixture
via pipette. The mixture was stirred at room temperature for
4 h, then filtered, and the solvent removed to yield 7 as a
brownish-red solid that was dried under vacuum; yield:
150 mg (>99%); mp 1858C (dec.). 1H NMR (DMSO-d6): d¼
1.99 (s, 3H), 2.95 (s, 12H), 4.6 (s, 4H), 6.78 (dd, 4H), 7.20 (s,
2H), 7.66 (dd, 4H), 9.83 (s, 1H); 13C NMR (DMSO-d6): d¼
24.6, 40.4, 52.1, 113.2, 114.2, 115.8, 134.4, 137.5, 150.5, 152.2,
168.9. HRMS: the mass spectrum was obtained from a dilute
acetic acid solution and compound was detected as an acetate
complex; [MþH2OþHþ]; calcd. for C28H36O4N3S2Pd:
648.1182; found: 648.1085.
Acknowledgements
Supportof this work by the National Science Foundation (CHE-
0010103), the Robert A. Welch Foundation and ArQule is grate-
fully acknowledged.
References and Notes
[1] M. E. van der Boom, D. Milstein, Chem. Rev. 2003, 103,
1759–1792.
Palladation of N-Acetyl-3,5-bis((p-
acetamidophenyl)thiomethyl)aniline
[2] M. Albrecht, G. van Koten, Angew. Chem. Int. Ed. 2001,
40, 3750–3781.
[3] C. M. Jensen, Chem. Commun. 1999, 2443–2449.
[4] I. Goettker-Schnetmann, M. Brookhart, J. Am. Chem.
Soc. 2004, 126, 9330–9338.
[5] J. T. Singleton, Tetrahedron 2003, 59, 1837–1857.
[6] a) H.-J van Manen, R. H. Fokkens, N. M. M. Nibbering,
F. C. J. M. van Veggel, D. N. Reinhoudt, J. Org. Chem.
2001, 66, 4643–4650; J. M. Pollino, L. P. Stubbs, M.
Weck, J. Am. Chem. Soc. 2004, 126, 563–567; c) C. S.
Consorti, G. Ebeling, F. Rodembusch, V. Stefani, P. R.
Livotto, F. Rominger, F. H. Quina, C. Yihwa, J. Dupont,
Inorg. Chem. 2004, 43, 530–536; d) A. Friggeri A, H.-J.
van Manen, T. Auletta T, X.-M. Li, S. Zapotoczny, H.
Schçnherr, G. J. Vancso, J. Huskens, F. C. J. M. van Veg-
To a solution of bis(acetamido) SCS ligand prepared above
(200 mg, 0.406 mmol) in 5 mL of DMF was added a solution
of Pd(TFA)2 (135 mg, 0.406 mmol) in 5 mL of DMF. The solu-
tion was stirred at room temperature for 5 h. The solvent was
removed under reduced pressure to give 8 as a maroon solid;
1
yield: 282 mg (97%); mp 1558C (dec.). H NMR (DMSO-d6):
d¼1.99 (s, 3H), 2.05 (s, 6H), 4.7 (s, 4H), 7.20 (s, 2H), 7.64
(dd, 4H), 7.79 (dd, 4H), 9.80(s, 2H), 10.20 (s, 1H); 13C NMR
(DMSO-d6): d¼23.9, 24.0, 49.6, 113.6, 119.7, 124.0, 132.7,
136.8, 141.2, 148.9, 168.1, 168.7; IR (KBr): n¼1675, 1592,
1261 cmꢀ1; HRMS: a mass spectrum was obtained from a di-
lute acetic acid solution and compound was detected with an
acetate ligand; [MþHþ þH2O] calcd. for C28H32O6N3S2Pd:
676.08168; found: 676.0551.
182
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 172–184

Mechanistic Studies of SCS-Pd Complexes Used in Heck Catalysis
FULL PAPERS
gel, D. N. Reinhoudt, J. Am. Chem. Soc. 2001, 123, 6388–
6395.
[20] K. Yu, W. Sommer, M. Weck and C. W. Jones, Adv.
Synth. Catal. 2005, 347, 161–171, the preceding paper
in this issue. We thank Professors Weck and Jones for
providing us with a copy of this manuscript before we
submitted this paper and for discussions about their
work.
[21] M. R. Eberhard, Org. Lett. 2004, 6, 2125–2128.
[22] K. Yu, W. Sommer, M. Weck, C. W. Jones, J. Catal. 2004,
226, 101–110.
[23] A. H. M. de Vries, J. M. C. A. Mulders, J. H. M. Momm-
ers, H. J. W. Henderickx, J. G. de Vries, Org. Lett. 2003,
5, 3285–3288.
[24] C. Rocaboy, J. A. Gladysz, Org. Lett. 2002, 4, 1993–1996.
[25] M. T. Reetz, E. Westermann, Angew. Chem. Int. Ed.
2000, 39, 165–168.
[26] A. S. Gruber, D. Pozebon, A. L. Monteiro, J. Dupont,
Tetrahedron Lett. 2001, 42, 7345–7348.
[27] V. P. W. Bohm, W. A. Herrmann, Chem. Eur. J. 2001, 7,
4191–4197.
[28] A. Sundermann, O. Uzan, J. M. L. Martin, Chem. Eur. J.
2001, 7, 1703–1711.
[29] C. Amatore, A. Jutand, J. Organometal. Chem. 1999, 576,
254–278.
[30] H. B. Kraatz, M. E. van der Boom, Y. Ben-David, D.
Milstein, Isr. J. Chem. 2001, 41, 163–171.
[31] B. L. Shaw, New J. Chem. 1998, 77–79.
[32] B. L. Shaw, S. D. Perera, E. Staley, E. A. Chem. Com-
mun. 1998, 1361–1362.
[33] T. Rosner, J. Le Bars, A. Pfaltz, D. G. Blackmond, J. Am.
Chem. Soc. 2001, 123, 1848–1855.
[34] T. Rosner, A. Pfaltz, D. G. Blackmond, J. Am. Chem.
Soc. 2001, 123, 4621–4622.
[35] A. J. Canty, Acc. Chem. Res. 1992, 25, 83–90.
[36] W. A. Herrmann, C. Brossmer, K. Oefele, C.-P. Reising-
er, T. Priermeier, M. Beller, H. Fischer, Angew. Chem.
Int. Ed. Engl. 1995, 34, 1844–1847.
[7] a) V. V. Dunina, O. N. Russ. Chem. Rev. 2004, 73, 309–
350; b) R. B. Bedford, Chem. Commun. 2003, 1787–179
6; c) J. Dupont, M. Pfeffer, J. Spencer, Eur. J. Inorg.
Chem. 2001, 1917–1927; d) N. J. Whitcombe, K. K. Hii,
S. E. Gibson, Tetrahedron 2001, 57, 7449–7476; e) J. G.
de Vries, Can. J. Chem. 2001, 79, 1086–1092. f) I. P. Be-
letskaya, A. V. Cheprakov, Chem. Rev. 2000, 100, 3009
–3066.
[8] J. Errington, W. S. McDonald, B. L. Shaw, J. Chem. Soc.
Dalton Trans. 1980, 2312–2314.
[9] a) M. Q. Slagt, G. Rodriguez, M. M. P. Grutters,
R. J. M. K. Gebbink, W. Klopper, L. W. Jenneskens, M.
Lutz, A. L. Spek, G. van Koten, Chem. Eur. J. 2004, 10,
1331–1344; b) J. S. Fossey, C. J. Richards, Christopher J.
Tetrahedron Lett. 2003, 44, 8773–8776; c) G. Jung, II,
S. U. Son, K. H. Park, K.-C. Chung, J. W. Lee, Y. K.
Chung, Organometallics 2003, 22, 4715–4720;
d) B. M. J. M. Suijkerbuijk, L. Shu, R. J. M. K. Gebbink,
A. D. Schlueter, G. van Koten, Organometallics 2003,
22, 4175–4177; e) H. P. Dijkstra, N. Ronde, G. P. M. van -
Klink, D. Vogt, G. van Koten, Adv. Synth. Catal. 2003,
345, 364–369; f) E. Diez-Barra, J. Guerra, I. Lopez-Sol-
era, S. Merino, J. Rodriguez-Lopez, P. Sanehez-Verdu,
J. Tejeda, Organometallics 2003, 22, 541–547.
[10] D. E. Bergbreiter, P. L. Osburn, Y.-S. Liu, J. Am. Chem.
Soc. 1999, 121, 9531–9538; b) D. P. Curran, K. Fischer, G.
Moura-Letts, Synlett 2004, 1379–1382; c) A. S. Gruber,
D. Zim, G. Ebeling, A. L. Monteiro, J. Dupont, Org.
Lett. 2000, 2, 1287–1290; d) D. Zim, A. S. Gruber, G. Eb-
eling, J. Dupont, A. L. Monteiro, Org. Lett. 2000, 2,
2881–2884; e) R. Gimenez, T. M. Swager, J. Mol. Catal.
A: Chem. 2001, 166, 265–273; f) H. Nakai, S. Ogo, Y. Wa-
tanabe Organometallics 2002, 21, 1674–1678.
[11] a) M. Ohff, A. Ohff, Andreas; M. E. van der Boom, D.
Milstein, J. Am. Chem. Soc. 1997, 119, 11687–11688;
b) A. Chanthateyanonth, H. Alper, J. Mol. Catal. A:
Chem. 2003, 201, 23–31; c) D. Morales-Morales, R. Re-
don, C. Yung, C. M. Jensen, Chem. Commun. 2000,
1619–1620; d) N. Solin, J. Kjellgren, K. J. Szabo, J. Am.
Chem. Soc. 2004, 126, 7026–7033; e) K. Kiewel, Y.-S.
Liu, D. E. Bergbreiter, G. A. Sulikowski, Tetrahedron
Lett., 1999, 40, 8945–8948.
[12] C. S. Consorti, G. Ebeling, F. Fabricio. F. Rominger, J.
Dupont, Adv. Synth. Catal. 2004, 346, 617–624.
[13] D. E. Bergbreiter, Chem. Rev. 2002, 102, 3345–3383.
[14] D. E. Bergbreiter, J. Li, Top. Curr. Chem. 2004 , 242 , in
press.
[15] D. E. Bergbreiter, P. L. Osburn, A. Wilson, E. M. Sink, J.
Am. Chem. Soc. 2000, 122, 9058–9064.
[16] D. E. Bergbreiter, P. L. Osburn, J. D. Frels, J. Am. Chem.
Soc. 2001, 123, 11105–11106.
[17] D. E. Bergbreiter, S. Furyk, Green Chem. 2004, 6, 280–
285.
[18] D. E. Bergbreiter, J. Li, Chem. Commun. 2004, 42–43.
[19] J. D. Frels, PhD dissertation, Texas A&M University,
USA, 2002.
[37] M. Beller, H. Fischer, W. A. Herrmann, K. Oefele, C.
Brossmer, Angew. Chem. Int. Ed. Engl. 1995, 34, 1848–
1849.
[38] J. A. Loch, M. Albrecht, E. Peris, J. Mata, J. W. Faller,
R. H. Crabtree, Organometallics 2002, 21, 700–706.
[39] H. R. Rogers, C. L. Hill, Y. Fujiwara, R. J. Rogers, H. L.
Mitchell, G. M. Whitesides, J. Am. Chem. Soc. 1980, 102,
217–226.
[40] M. Portnoy, D. Milstein, Organometallics 1993, 12, 1665–
1673.
[41] W. A. Herrmann, V. P. W. Bohm, C.-P. Reisinger, J. Or-
ganomet. Chem. 1999, 576, 23–41.
[42] J. Dupont, A. S. Gruber, G. Fonseca, A. L. Monteiro, G.
Ebeling, R. A. Burrow, Organometallics 2001, 20, 171–
176.
[43] A. F. Littke, G. C. Fu, Angew. Chem. Int. Ed. 2002, 41,
4176–4211.
[44] P. Dani, M. Albrecht, G. P. M. van Klink, G. van Koten,
Organometallics 2000, 19, 4468–4476.
[45] D. E. Bergbreiter, Y.-S. Liu, P. L. Osburn, J. Am. Chem.
Soc. 1998, 120, 4250–4251.
[46] D. E. Bergbreiter, S. D. Sung, J. Li, D. Ortiz, P. N. Ham-
ilton, Org. Proc. Res. Development 2004, 8, 461–468.
Adv. Synth. Catal. 2005, 347, 172–184
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
183

David E. Bergbreiter et al.
FULL PAPERS
[47] D. E. Bergbreiter, R. Hughes, J. Besinaiz, C. Li, P. L. Os-
burn, J. Am. Chem. Soc. 2003, 125, 8244–8249.
[48] D. R. Anton, R. H. Crabtree, Organometallics 1983, 2,
855–859.
D. W. Brown, E. M. Staudt, Organometallics 1985, 4,
1819–1830.
[50] J. A. Widegren, R. G. Finke, J. Mol. Catal. A: Chem.
2003, 198, 317–341.
[49] G. M. Whitesides, M. Hackett, R. L. Brainard, J.-P. P. M.
Lavalleye, A. F. Sowinski, A. N. Izumi, S. S. Moore,
184
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 172–184